is 28 miles to the north-east of Sahankat in a direct line measured on the map, or about 35 miles by road. The distance is therefore only 5 yojanas, instead of 12, as noted by Fa-Hian. It cannot be placed further to the north-east without increasing its distance from Banâras, and lessening its distance from Vaisâli. Now the former is limited by Hwen Thsang to 700 li, or 117 miles, and the latter is fixed by Fa-Hian himself at 25 yojanas, or 175 miles; and as both estimates agree very closely with the actual position of Kasia, I am satisfied that Fa-Hian's 12 yojanas must be a mistake. Anrudhwa, near Kasia, is exactly 111 miles to the north-north-east of Banâras, measured in a direct line on the map, and cannot, therefore, be less than 120 miles by road. The distance between Kasia and Vaisâli, by the route which I marched, is just 140 miles; but this was along the new straight lines which have been laid out by the British authorities. By the old winding native tracks the distance would have been much greater, or certainly not less than 160 miles.

At the time of Hwen Thsang's visit the walls of Kusinagara were in ruins, and the place was almost deserted; but the brick foundations of the old capital occupied a circuit of about 12 li, or 2 miles. The existing ruins between Anrudhwa and Kasia are scattered over a much larger space; but some of these were certainly outside the city, and it is now quite impossible to ascertain its exact limits. It most probably occupied the site of the mound of ruins to the north-east of the village of Anrudhwa. The spot where Buddha obtained Nirvána would then correspond with the site of the stupa and ruins now called

the south of Kahaon, and 7 miles below the confluence of the two rivers. From Kasia to the Mahili Ghât the route would have passed through the ancient towns of Khukhundo and Kahaon, both of which still possess many remains of antiquity. But the former is only 28 miles from Kasia, while the latter is 35 miles. Both are undoubtedly Brahmanical; but while the ruins at Khukhundo are nearly all of middle age, those at Kahaon are at least as old as the time of Skanda Gupta, who lived several centuries before the time of Hwen Thsang. I am inclined, therefore, to prefer the claim of Kahaon as the representative of Hwen Thsang's ancient city, partly on account of its undoubted antiquity, and partly because its distance from Kasia agrees better with the pilgrim's estimate than that of the larger town of Khukhundo.*

Páwá, or Padraona.

In the Ceylonese chronicles the town of Páwá is mentioned as the last halting-place of Buddha before reaching Kusinagara, where he died. After his death it is again mentioned in the account of Kâsyapa's journey to Kusinagara to attend at the cremation of Buddha's corpse. Pawa was also famous as one of the eight cities which obtained a share of the relics of Buddha. In the Ceylonese chronicles it is noted as being only 12 miles from Kusinagara,† towards the Gandak river. Now 12 miles to the north-north-east of Kasia there is a considerable village named Padaraona Padara-vana, with a large mound covered with broken bricks, in which several statues of *See Map No. XI. for the positions of both places.

Turnour, Journ. Asiat. Soc. Bengal, viii. 1005; note from B dhaghoso.

yet I can readily believe that its more popular name amongst all classes would have been the "statue of the Dead Prince." I am therefore of opinion that the name of Máthá Kuär, which still clings to the ruins of Kasia, has a direct reference to the death of Buddha, which, according to his followers, took place at Kusinagara, on the full moon of the Vaisâkh, 543 B.C. The continuance of this name down to the present day is a strong argument in favour of the identification of Kasia as the "death-place" of Buddha.

Khukhundo-Kahaon.

On leaving Kusinagara, Hwen Thsang directed his steps towards Banâras, and after having travelled about 200 li, or 33 miles, to the south-west, he reached a large town where lived a Brahman who was devoted to Buddhism.* If we adhere rigidly to the south-west bearing, we must identify this large town with Sahankat, near Rudrapur. But this place has already been identified with Pippalavana, and is not upon the high-road to Banâras. As Hwen Thsang specially mentions the Brahman's hospitality to travellers going and coming, it is certain that the large town must have been on the high-road between Kusinagara and Banâras. Now the high-road could never have passed through Rudrapur, as it would have entailed the passage of the Rapti in addition to that of the Ghagra, while Rudrapur itself is not on the direct line to Banâras. It is quite clear that the high-road must have crossed the Ghagra somewhere below the junction of the Rapti. According to the people, the old passage of the Ghagra was at Mahili, 4 miles to *Julien's Hiouen Thsang,' ii. 349:

ur, which is on the Ganges just 50 miles to the east of Banâras. The present name was given. by the Muhammadans, and is said to be only a slight alteraion of the original Hindu name of Garjpur. This is most probably the name referred to by Hwen Thsang, as Garjan, which means primarily any roaring noise, signifies also "battle," and Garjana-pati is a title of the "god of war." Ghâzipur is now a large city about 2 miles in length, and 5 or 6 miles in circumference. Hwen Thsang estimates the circuit of the district at 2000 li, or 333 miles, which is almost exactly the size of the tract lying between the Ghagra on the north and the Gomati on the south, from Tanda on the west to the confluence of the Ganges and Phagra.

At 200 li, or 33 miles, to the east of the capital, ven Thsang visited the Aviddhakarna monastery, h was adorned with very fine sculptures. Followhe bearing and distance, this place should be for in the neighbourhood of Baliya on the bank anges. Aviddhakarna means the "pierced I think it possible that the name may still in Bikapur, a village 1 mile to the east Aviddhakarna-pura might easily be npur and Bikanpur. It seems prohe same place that is menme of the "Vast soli